The present invention relates in general to data storage systems. More particularly, the present invention relates to an apparatus and method for reclaiming a disk substrate for use in a data storage device.
A typical data storage device includes a medium for storing data, typically in magnetic, magneto-optical or optical form, and a transducer used to write and read data respectively to and from the medium. A disk drive data storage device, for example, includes one or more data storage disks coaxially mounted on a hub of a spindle motor. The spindle motor rotates the data storage disks at speeds typically on the order of several thousand or more revolutions-per-minute. Digital information, representing various types of data, is typically written to and read from the data storage disks by one or more transducers, or read/write heads, which are mounted to an actuator assembly and passed over the surface of the rapidly rotating disks.
In a typical magnetic disk drive, for example, data is stored on a magnetic layer coated on a disk substrate. Several characteristics of disk substrates significantly affect the areal density of a disk drive. One such characteristic that significantly affects the areal density of a disk drive is the uniformity of the surface of the disk substrate, i.e., the absence of substrate surface defects. It is generally recognized that minimizing the flyheight, i.e., the clearance distance between the read/write head and the surface of a data storage disk, generally provides for increased areal densities. It is also recognized in the art, however, that the smoothness of the surface of a data storage disk becomes a critical factor and design constraint when attempting to minimize the flyheight. A significant decrease in flyheight provided by the use of data storage disks having highly uniform recording surfaces can advantageously result in increased transducer readback sensitivity and increased areal density of the disk drive. The uniformity of disk substrate surfaces affects the uniformity of the recording surfaces because the layers sputtered onto the disk substrate, such as the magnetic layer, replicate any irregular surface morphology of the disk substrate.
Conventionally, disk substrates have been based upon aluminum, such as NiP coated Al/Mg alloy substrates. Coating the aluminum magnesium alloy with a nickel-phosphorus plate provides a harder exterior surface which allows the disk substrate to be polished and superfinished. A conventional superfinishing process and slurry is described in U.S. Pat. No. 6,236,542 to Hartog et al., which is assigned to the assignee of the present application. Typically, the Al/Mgxe2x80x94NiP substrate is superfinished to a smooth finish with a colloidal slurry, e.g., a pH adjusted aqueous slurry containing colloidal silica and/or colloidal alumina particles and an etching agent such as aluminum nitrate, prior to sputtering with thin film magnetic coatings. The colloidal alumina and silica slurries are then cleaned from the substrate by the general cleaning mechanisms of mechanical scrubbing, dispersion and etching. Surfactants and pH are generally used for dispersion cleaning, where the surfactant and pH act to separate the slurry particles from each other and from the substrate. Etching is generally accomplished by acids and acid soaps that erode or dissolve the substrate material beneath embedded slurry particles (under-cut) to release them from the substrate. Typical acids in use for NiP plated Al-based substrates include, for example, straight phosphoric acid, nitric acid, hydrofluoric acid-based soaps and phosphoric acid-based soaps. The straight acids generally have a pH less than 1 and the soaps generally have pH""s above 1.
After cleaning, the substrates are sputtered with a series of layers, e.g., a chrome underlayer, a magnetic layer and a carbon protection layer. If residual slurry particles are left on the substrate or if there is galling to the relatively soft NiP layer, the sputtered layers replicate the irregular surface morphology, creating a bumpy surface on the finished disk. When the read/write head glides over the surface, it crashes into bumps created by the residual particles and/or damage that is higher than the glide clearance. This is known as a glide defect, which can ultimately cause disk drive failure. These bumps further cause magnetic defects, corrosion and decreased disk life. Thus, the residual slurry particles and/or damage needs to be removed from the superfinished substrate surface so that the substrate is as smooth as possible.
Unfortunately, aluminum-based substrates have relatively low specific stiffness, as well as relatively low impact and dent resistance. For example, the relatively low specific stiffness of the Al/Mgxe2x80x94NiP substrates (typically 3.8 Mpsi/gm/cc) makes this type of disk substrate susceptible to environmental forces which create disk flutter and vibration and which may cause the read/write head to impact and dent the disk substrate surface.
More recently, glass substrates have been used for disk drives in portable devices, such as laptop computers. Glass substrates have a higher impact and dent resistance than aluminum-based substrates, which is important in portable devices where the unit is subject to being bumped, dropped and banged around, causing the read/write head to bang on the disk substrate surface. Moreover, the specific stiffness of glass or glass-ceramic substrates (typically xe2x89xa66 or 7 Mpsi/gm/cc) is typically higher than that of aluminum-based substrates.
An additional benefit of glass is that it is easier to polish to and maintain as a smooth surface finish (as compared to NiP) than aluminum-based substrates. A smoother substrate allows the read/write head to fly closer to the disk, which produces a higher density recording. Glide height for some computer disk drives is on the order of 20 nanometers (about 200 xc3x85) and less, which is an extremely small interface distance. Thus, the fact that glass substrates can be polished to smoother finishes makes an industry shift from Al-based substrate s to glass substrates desirable, not only for disk drives used in portable devices, but for disk drives used in stationary devices as well.
The surface uniformity of glass substrates can still present a problem, however, especially for low glide heights (typically xe2x89xa620 nanometers) and near contact recording. Just as with aluminum-based substrates, the surface of the glass substrate needs to be polished and superfinished with a slurry to provide an atomically smooth surface prior to sputtering. Such a conventional superfinishing polish process and slurry is also described in the above referenced U.S. Pat. No. 6,236,542 to Hartog et al. Typically, the glass substrate is superfinished to a smooth finish with a colloidal slurry, e.g., a pH adjusted aqueous slurry containing colloidal silica and/or colloidal alumina particles and an etching agent such as cerium sulfate, prior to sputtering with thin film magnetic coatings.
In this conventional superfinishing polish process colloidal silica particles attach to the surface being polished not only by the usual London dispersion forces, van der Waals forces and hydrogen bonding, but unlike NiP, also by molecular bonding even though the slurry has the usual stabilizing agents used in the colloidal silica to prevent the silica particles from sticking to each other (interparticle siloxane bonding), charge repulsion and/or steric stabilizers. Standard methods of scrubbing with soaps using polyvinyl alcohol (PVA) pads, ultrasonics or megasonics will not remove any significant percentage of such molecular bonded silica particles. Just as with aluminum-based substrates, if these particles are left in place on the glass substrate, glide defects occur that can ultimately cause disk drive failure. These glide defects further cause magnetic defects, corrosion and decreased disk life.
A less-than-optimal solution to this problem is to use stronger acid or base solutions than the cleaning soap, to etch the glass substrate or undercut the slurry particles similar to what can be done to remove hard alpha alumina from Al/Mgxe2x80x94NiP substrates after non-superfinish polish slurries. The surface finish of glass and NiP substrates are, however, damaged by such a technique by surface topography change such as pitting and chemical composition changes. Glass has low resistance to acid etching and overly aggressive acid solutions, such as hydrofluoric acid and caustic etching at high pH""s and temperatures. Damage and compositional change to the superfinished glass surface will adversely affect the morphology of layers deposited by subsequent sputtering processes and can cause magnetic, glide and corrosion failures.
A better solution to this problem is to use a cleaning polish etch solution/process (a process performed by running disk substrates on a polishing pad using an etch solution instead of a slurry, i.e., there are no slurry particles in the cleaning polish etch solution) with acid, neutral or base solutions to etch the glass substrate and/or the attached slurry particles under polish conditions thereby maintaining the superfinish surface while removing the superfinish polish slurry debris by etching and dilution. Such a cleaning polish etch solution/process is disclosed in the copending application Ser. No. 09/976,412 entitled xe2x80x9cCLEANING POLISH ETCH COMPOSITION AND PROCESS FOR A SUPERFFNISHED SURFACE OF A SUBSTRATExe2x80x9d, assigned to the same assignee as the present application. Etching by itself (i.e., the first solution discussed above) with PVA scrub, ultrasonics or megasonics is what has been done to remove slurry particles from Al/Mgxe2x80x94NiP or glass substrates, but with the less than 20 nm glide heights now in use, a cleaning polish etch solution/process ensures 100% surface cleaning of particles that small (i.e., the lower the glide height, the smaller the particles needing to be removed, and thus the more difficult they are to remove) while maintaining the surface finish. The cleaning polish etch process, however, adds equipment and handling costs. Nonetheless, without the cleaning polish etch process the surface of the glass substrate can be damaged by using only chemical etch due to the low resistance of the glass material to acid etching or overly aggressive caustic etch solutions.
An even better solution to this problem is to use a self-cleaning colloidal slurry and process, such as disclosed in the copending application Ser. No. 09/976,167 entitled xe2x80x9cSELF-CLEANING COLLOIDAL SLURRY COMPOSITION AND PROCESS FOR FINISHING A SURFACE OF A SUBSTRATExe2x80x9d, assigned to the same assignee as the present application. The slurry comprises a carrying fluid, colloidal particles, etchant, and a surfactant adsorbed and/or precipitated onto a surface of the colloidal particles and/or substrate. The surfactant has a hydrophobic section that forms a steric hindrance barrier and substantially prevents contaminates, including colloidal particles, from bonding to the substrate surface. Subsequent cleaning with standard soap solutions removes substantially all remaining contamination from the substrate surface.
After cleaning, the glass substrate is typically subjected to chemical strengthening. Chemical strengthening is known in the art of treating glass. In chemical strengthening, the substrate is immersed in molten potassium nitrate and/or sodium nitrate for anywhere from 1 to 8 hours to strengthen the glass against breaking.
Effective techniques for polishing, cleaning and chemical strengthening can provide low glide yield losses and decreased disk drive failures. Nonetheless, because of the extremely large number of glass disks produced, even low glide yield losses and customer return rates dictate that a large number of glass disks are, or ultimately will be, rejected for various reasons including glide height, magnetic defects or disk drive failure. The inventors are not aware of any manufacturer in the disk drive industry that reclaims rejected glass disks. Huge quantities of rejected glass disks are merely discarded, adding to production costs and waste disposal demands.
If the market trend toward glass substrates in disk drives is to flourish, a mechanism for reclaiming rejected glass disks is required. Preferably, such a mechanism would improve production costs (as compared to discarding the rejected glass disk and making a new glass disk from a blank) and provide high quality glass disks.
An object of the present invention is to provide a mechanism of reclaiming a disk substrate for use in a data storage device.
Another object of the present invention is to provide a mechanism of reclaiming a disk substrate for use in a data storage device that improves production costs (as compared to discarding a rejected disk and making a new substrate from a blank) and provides a high quality disk.
These and other objects of the present invention are achieved by an apparatus and method of reclaiming a disk substrate for use in a data storage device. According to one aspect of the present invention, a disk substrate for use in a data storage device is reclaimed by stripping a data storage disk, polishing the stripped disk and cleaning the polished disk. According to another aspect of the present invention, an apparatus for polishing a disk substrate for use in a data storage device precisely controls the relative velocity of polishing pads as seen by the substrate so that a substantially equal amount of stock is removed from each side of the substrate.
In an exemplary embodiment, at least a portion of a layer (e.g., a sputtered magnetic recording layer) of a data storage disk is stripped by, for example, immersing the disk in an acid/oxidizing bath comprising water, hydrochloric acid and nitric acid. The stripped disk is polished (and preferably superfinished) in a carrier between an upper polishing pad and a lower polishing pad, with the relative velocity of the polishing pads as seen by the disk being precisely controlled so that an equal amount of stock is removed from each side of the disk. Because of the high level of stress in a chemically strengthened disk substrate, a side-to-side difference in the amount of stock removed will cause the disk substrate to have high curvature or out-of-flatness (a condition that renders the disk substrate useless and is often referred to as xe2x80x9cconingxe2x80x9d). Preferably, several stripped disks are sorted into groups based on disk thickness, and disks from one of the groups are simultaneously polished in the carrier. Sorting improves stock removal uniformity from disk to disk. The polished disks are cleaned and, preferably, ordered in a cassette for stacking in the data storage device based on disk thickness to more easily meet a mean center specification. A data storage disk for use in the data storage device may be provided by applying a recording layer over the surface of the polished disk.
The cost of reclaiming a disk substrate can be lower, and the quality higher, than making a new one from a blank. For example, the reclaimed disk substrate need not be subjected to a subsequent chemical strengthening process and an additional cleaning process thereafter. That is, the substrate of the rejected disk was already subjected to a chemical strengthening process prior to rejection of the disk and the chemical strengthening remains even after the disk substrate is recycled. In addition to reducing production costs, the elimination of subsequent chemical strengthening and cleaning processes means that the polished surface of the reclaimed disk substrate will not be degraded by adverse surface effects (e.g., pitting and asperities) that can be caused by the chemical strengthening process.